This project has two purposes, (a) to examine the validity of a response-to-intervention (RTI) approach to identifying children with reading disabilities (RD) who are beyond the initial stages of reading acquisition and (b) to describe the cognitive and neurobiological characteristics of children who respond differentially to different intensities of intervention. The emotional and economic sequelae of reading failure in a highly literate society represent a public health concern. Reading research with at-risk and disabled readers has focused on the primary school grades with little attention to children who continue to struggle with reading or experience difficulties for the first time after second grade. These children are the focus of the proposed investigation.Three research questions are designed to extend understanding of post-primary RD: (a) what combination of measures and procedures provide an effective screen to identify upper elementary children for secondary reading prevention within an RTI framework?, (b) does the probability of reading failure defined by the screening battery decrease as a function of secondary prevention both immediately and longitudinally?, and (c) what are the cognitive and neurobiological characteristics of at-risk children who respond differentially to secondary and tertiary prevention? A novel aspect of the design is defining risk of reading failure continuously as a probability level rather than as a dichotomy. The research plan incorporates a multiple cohort, longitudinal design in which three successive cohorts of third and fourth grade children are selected (N = 750). In Year 1, measures tested for the identification battery include word and text-level reading measures, oral language, and attention. Measures of literacy growth also are used to determine if prediction precision is improved. Multiple regression and dominance analysis assess impact on reading measured continuously and to reduce the variable set. Logistic regression is used to develop the best fitting algorithm to predict reading failure and produce probabilities. This algorithm is applied in Years 2 and 3 to identify children with higher probabilities of failure (N=140). These children are randomly assigned to either a no treatment control or a 12 week intervention. At-risk children in the Year 3 cohort who do not respond to the intervention will be randomly assigned to tertiary prevention or a wait list control. These children participate in experiements designed to elaborate the neurobiological signature of post-primary RD.